Paul Tudor Jones Just Issued a Terrifying Warning for 2027

Anthropic

Across the shows: 17 speakers in other episodes, 10 bullish and 25 bearish calls, past 30 days

• Jeff Snider · Atlas FinancialBearish

  Anthropic's losses and funding commitments undercut the idea that AI is funded by profitable companies.

  Snider contrasts this AI cycle with the dot-com era, arguing it was supposed to be different because profitable companies were investing in AI. He cites Anthropic's disclosed figures, showing revenue far below its losses and very large funding commitments, and says that premise is turning out to be less true. He argues that as these numbers grow, the credit market gets more information and must borrow more to fund them. The transcript's figures are garbled, so only the direction of the losses and funding needs is clear.

  “Anthropic, you know, numbers that came out with them, billion in revenue, not even 5,000,000,000 in revenue, 40,000,000,000 in losses was $1,112,000,000,000 in operating losses and a 500,000,000,000 in funding commitments.”at 11:03

US stock market outlook

• Paul Tudor Jones · Tudor Investment Corporation

  The market is a two-speed split between AI adopters and non-adopters moving in opposite directions, like 1999.

  Tudor Jones says the market still feels like 1999. He describes a two-speed market that is completely bifurcated: companies adopting AI and companies that are not are going in opposite directions. He notes this was also the pattern in 1999, when the market split in a similar way before the dot-com bust.

  “Boy, it still feels like 1999, a two speed market. You've got it completely perfectly bifurcated in the sense that you have obviously everything that's adopting AI and everything that's not, and they're going in opposite directions.”at 1:13

• Paul Tudor Jones · Tudor Investment Corporation

  2027 will be very challenging for US stocks because equity flows are the most negative since 2009.

  Tudor Jones sums up buybacks, M&A and the other positive corporate flows that have powered the US market since about 2009, then subtracts IPOs, secondaries and lock-up expiries (unlocks). The net flow picture is the most negative he has seen since 2002 for the US market, and certainly the most negative since 2009. He says flows are what he watches most, and that they make 2027 very challenging for the stock market.

  “2027 is gonna be really, really challenging for the stock market. And I say that Because? Because if I just look at the flows... You probably got the most negative flow picture for The US stock market that we've had say since 2002... Certainly the most negative flow picture since 2009.”at 1:33

• Jeff Snider · Atlas Financial

  The market is AI-or-nothing, with breadth at dot-com-bust lows, which is risky if the leaders turn.

  Snider says the stock market is narrowly driven, essentially AI or nothing. Goldman Sachs' preferred breadth gauge is at levels not seen since the dot-com bust. Narrow leadership is not necessarily bad while the leaders keep leading, but at the tail end of a cycle like the dot-com era it becomes a last gasp of everyone piling in. He notes AI no longer looks like a leader, while consumer discretionary and staples are seeing defensive rotation. The lack of breadth makes the market more dangerous if the leaders turn.

  “It's basically AI or nothing. In fact, market breadth is down to levels that according to certain sources like Goldman Sachs, their preferred market breadth gauge haven't seen anything like this since the dot com bust.”at 4:07

Credit markets

• Jeff Snider · Atlas Financial

  The 1987 crash and 1990 recession came from a credit-cycle downswing, not interest rates.

  Snider says the 1987 crash happened during a credit-cycle downswing after a major boom in lending, especially by non-bank shadow lenders in the mid-1980s. The S&L crisis was heating up by 1987, and it drove the 1990 to 1991 recession, with the Iraq-Kuwait oil shock as a trigger. He argues people focus on interest rates when they should focus on the credit cycle, since a shift toward risk aversion in the banking and lending system is what undermines lofty stock valuations. He says today's markets show a similar pattern.

  “Crash of eighty seven took place not accidentally, not coincidentally during the downswing in the credit cycle.”at 15:24

• Jeff Snider · Atlas Financial

  Triple-C spreads are above their 2022 peak and rising steadily, signalling a credit-cycle downswing.

  Snider says triple-C spreads are at about 1,260 basis points, above the peak of the 2022 credit difficulties. Spreads have risen steadily since the Tricolor and First Brand failures, unlike the sharp spike-and-reversal of August 2024 or the tariff shock of spring 2025. He says the market is moving from a flashing yellow light to a flashing pink light as thresholds are crossed. He expects serious trouble around 1,500 to 1,600 basis points in triple-C. He also notes investment-grade spreads track the VIX, so broader credit stress would reach stocks.

  “the triple Cs now, triple Cs at [1,260] basis points at least for this index. Haven't seen anything like that since really the height of the credit market difficulties in 2022.”at 18:53

US growth & recession risk

• Jeff Snider · Atlas Financial

  The two-speed economy is really a sequential slowdown, and the lower part will spread to the top and markets.

  Snider disputes the K-shaped framing. He says one part of the economy, the bottom, slows down first while the top keeps performing as if nothing is wrong, supported by a wealth effect. Once the slower segment slows enough, it hits the upper part of the economy and the marketplace too, so everyone ends up in the slow lane. He argues this would be a significant risk to the stock market.

  “The two speed economy, they call it the k shaped economy. It's not really. Because what ends up happening is one part of the economy slows down first.”at 3:16

AI bubble & tech valuations

• Jeff Snider · Atlas Financial

  AI is a typical bubble: extreme expectations, investors betting without knowing who will win.

  Snider says AI is a disruptive technology with two possible tails, so nobody knows who will win or lose. In every bubble, people throw money at everything hoping to pick a winner, and set expectations that cannot be met. He argues the Broadcom CEO's comments reinforce that this time is not different. He adds that adoption questions mean growth may be parabolic but less so than current expectations, which would cause trouble. He says AI will be a wonderful technology eventually, but the 2030s must climb over 2027 first.

  “We know this is a bubble. We know that this is a typical bubble, typical cycle that goes along with disruptive technology and technological adoption.”at 7:16

AI infrastructure & data centers

• Jeff Snider · Atlas Financial

  The AI buildout is debt-fueled, and credit markets, not equity, will be the biggest constraint on it.

  Snider argues the AI trend depends on investors continuing to fund it with debt. He cites the Broadcom CEO saying capital will be the biggest constraint on the AI trend. He points to credit markets starting to balk at the larger numbers and to hand-overs of tens of billions in bridge loans and financing arrangements. In his view, debt-market willingness, rather than equity valuations or future earnings, will determine how far the AI trend can go.

  “It's a debt fueled bubble in every sense of the word. It is dependent upon investors continuing to shovel endless amounts of money at AI.”at 9:55

Private credit

• Jeff Snider · Atlas Financial

  Private credit is the epicenter of the credit downswing, like subprime mortgages before 2008.

  Snider says his team has focused on private credit for most of the year because that is where the credit cycle's epicenter has been. He draws a comparison to subprime mortgages before 2008, while stressing that he is not predicting a 2008 rerun. He notes that the credit cycle has not had a downswing since 2006 to 2007, and that these cycles follow similar patterns of human behaviour.

  “We've been focusing on private credit for the better part of the year because that's where the epicenter has been... I hesitate to use this analogy, but same way that subprime mortgages were before 2008.”at 17:24